Variability in ejection fraction calculations

In the last 8 years I have had many stress tests which were soon followed by Cardiac catheterization. On a number of occasions the stress test showed ejection fraction to range between 20-30%. However, a catheterization (angiogram) performed within  days or weeks showed much higher EF, ( 42-55%)

I would appreciate any explanation for this vast difference in numbers.

Thanks.

[nspower]

You have raised a very interesting question.
In my case it was other way round. My echo showed EF between 30-35% and catheterization (agiogram) within a week showed EF as 22%. I do not know the reason but over a period of time I have started doubting accuracy of measurement of ejection fraction. It looks like accuracy of EF is highly dependent on the person (and machine) who measures it.

I am also unable to establish any correlation between ejection fraction and a person's capability to exert himself/herself. My exertion capability has substantially increased in last two years after MI but the ejection fraction has remained static at 30-35%.

[kenkeith]

ACC/AHA Guidelines-Device Based Therapy for an Echo in low CHF.

"When performed properly , echo EF is more reliable than cath EF or SPECT nuclear EF. We also acknowledge that the determination of LVEF lacks a "gold standard" and that there may be variation among the commonly used clinical techniques of LVEF determination. All clinical methods of LVEF determination lack precision, and the accuracy of techniques varies among laboratories and institutions. On the basis of these considerations, the present writing committee recommends that clinicians use the LVEF determination that they believe is the most clinically accurate and appropriate in their institution. Anthony C. Pearson, MD, FACC

It seems the measurement by volume with an echo is more accurate than a cath.  When I was admitted in the hospital my EF was 13% with a cath and the echo 29%.  I was told if I had had an EF of 13% I wouldn't have been able to walk up the stairs. ...every echo with interpretable images of LV function should have EF measured and reported out precisely along with end-diastolic and end-systolic volumes.